NondeterministicBeliefTracker

class NondeterministicBeliefTracker[ValueType](

pomdp: stormpy.models.SparsePomdp[ValueType],

options: stormpy.pomdp.developer.NondeterministicBeliefTrackerOptions[ValueType] = ...,

)

Tracker for belief states (in sparse format) and uncontrollable actions

Template arguments:

• ValueType (type)

Available specializations:

• NondeterministicBeliefTracker[float]

• NondeterministicBeliefTracker[stormpy.numbers.Rational]

Construction

The specialization can be inferred from the first positional constructor argument (alternatively, the pomdp keyword argument).

Members

NondeterministicBeliefTracker.get_current_beliefs(

self: stormpy.pomdp.NondeterministicBeliefTracker[ValueType],

) → set[stormpy.pomdp.SparseBeliefState[ValueType]]

Current beliefs

NondeterministicBeliefTracker.get_current_dimension(

self: stormpy.pomdp.NondeterministicBeliefTracker[ValueType],

) → int

Dimension of the current beliefs

NondeterministicBeliefTracker.get_current_observation(

self: stormpy.pomdp.NondeterministicBeliefTracker[ValueType],

) → int

Last observation

NondeterministicBeliefTracker.get_current_risk(

self: stormpy.pomdp.NondeterministicBeliefTracker[ValueType],

max: bool = True,

) → ValueType

Maximal (or minimal) risk of the current beliefs

NondeterministicBeliefTracker.get_number_of_beliefs(

self: stormpy.pomdp.NondeterministicBeliefTracker[ValueType],

) → int

Number of current beliefs

NondeterministicBeliefTracker.has_timed_out(

self: stormpy.pomdp.NondeterministicBeliefTracker[ValueType],

) → bool

Whether the last reduction timed out

NondeterministicBeliefTracker.reduce(

self: stormpy.pomdp.NondeterministicBeliefTracker[ValueType],

) → int

Reduce the set of beliefs

NondeterministicBeliefTracker.reset(

self: stormpy.pomdp.NondeterministicBeliefTracker[ValueType],

observation: SupportsInt | SupportsIndex,

) → bool

Start tracking with the given initial observation

NondeterministicBeliefTracker.set_risk()

For NondeterministicBeliefTracker[float]:

set_risk(self: stormpy.pomdp.NondeterministicBeliefTracker[ValueType], risk: collections.abc.Sequence[typing.SupportsFloat | typing.SupportsIndex]) -> None

Set the risk of each state

For NondeterministicBeliefTracker[stormpy.numbers.Rational]:

set_risk(self: stormpy.pomdp.NondeterministicBeliefTracker[ValueType], risk: collections.abc.Sequence[ValueType]) -> None

Set the risk of each state

NondeterministicBeliefTracker.track(

self: stormpy.pomdp.NondeterministicBeliefTracker[ValueType],

new_observation: SupportsInt | SupportsIndex,

) → bool

Update the beliefs with the given observation